Formats of Rounds - II

Update (25 Dec 2006): See this post for an update on this subject. Some obsolete/incorrect portions have been struck off. In continuation to Formats of Rounds - I

Infinite Rebounds

Since Direct and Passing doesn't work in its usual form for reasons noted before, it has quickly been replaced by the Infinite Rebounds system as the de facto quizzing format in all the good quizzes. A caveat here is that of the qualifier "good quizzes", as a great many "popular" quizzes aren't aware of this system, or if they are, choose to shy away from employing it - an exploration later.

The commonly agreed conventions of Infinite Rebounds:

1. The choice of the team to receive the next question depends on how the previous question was answered. If the previous question was answered by a team, the next one is put to the team next in order from it. Alternately, if it went unanswered, the new question is simply advanced to the team next to the one that first received the last question.

   As an example, consider six teams A to F. Say a question starts with Team B. Turns are taken in order to attempt answers and Team E eventually answers correctly. In this case Team F would receive the next question. In case the question was left unanswered, Team C would receive first crack at the next question.

2. No question may be posed first to the same team twice in a row. (More elaboration on this later in this post.)

3. The beginning of a new round does not require resetting of the asking order i.e. unlike in the D&P method, each round does not start with the same team or does not have a predecided beginning point. An element of randomness appears in this system as it is not evident who would have answered the last question of the preceding round. This prevents a great deal of rigging (nothing can prevent quizmasters "choosing" their questions, of course, but it hampers rigging when the sequence of questions is already embedded in a visual presentation as is usually the case these days).

4. Each question has the same value for the correct answer.

5. No question in the rounds with using the Infinite Rebounds system has penalties i.e. no negative scoring for wrong answers.

According to me, the whole philosophy of Infinite Rebounds can be summed up with the sentence: Be as fair as possible to the participants. That is really the crux of the structure. If another system which is even better at being fair is possible, I don't have any doubts that it will replace Infinite Rebounds as we know it today.

I'd like to spend some more time further examining the implications of the above conventions. The point about the equal value and penalties aren't really a fundamental part of the structure of Infinite Rebounds, which is after all a set of rules for passing. However, they do conform to the spirit behind encouraging the use of this collection of rules - so they've been bundled up under the same banner. If someone so desired, a system of different values for direct and for passed questions may be employed, or points subtracted for guessing wildly thus inhibiting the choice to dare an answer, but that would bring what I like to think of as Instant Recoil (it can be seen on the faces of experienced quizzers on stage and looks as if they've swallowed a raccoon, and by the immediate broadcast of this feeling to fellow quizzers around).

The first rule presents the main algorithm of the format. It gets rid of one of the fundamental problems of the D&P method where a team could get a passed question correct and then a direct one for itself, boosting its tally without other teams having had a say in proceedings. Infinite Rebounds puts the team with the last correct answer at the end of the queue, forcing the movement of the "first attempt marker" to constantly circulate (the "first attempt marker" indicates how teams got first crack at a question). Additionally, this implicitly provides an element of randomness that is very useful in preventing many of the known methods of "black magic" that happens at quizzes :-). Teams also need not be very glum about having drawn seats next to so-called "studs" that in the D&P method would've eaten up much of the food supply of questions. Instead, such a team may be rewarded for its position, as it could potentially get many more first shots at questions. This one fact is enough to serve as an endorsement of this format over the conventional system. The clinching factor is that on most occasions, the team that has given the most correct answers wins. Of course, the number of correct answers is a factor of the number of attempts a team receives. There are cases where among two equally balanced teams slug it out and one team sneaks through because it got more questions to answer than its principal rival, but both teams will agree that it wasn't because the system did them in - the winners still had to provide more correct answers than the runners-up. Contrast this with the D&P team where quite often in a tight contest, the winning team could've had lesser attempts, lesser number of correct answers and still win.

The second point in the list, the one about not giving first attempt of a question to the same team twice in a row, requires some elucidation. Such a case may not be readily apparent. This can happen when (taking an example) say, Team B gets the next question in a round. They can't answer it and it passes normally till Team A (the last team in line for that question) answers it correctly. Now, by the normal passing rule which simply states that the team next in order to the one that got the last answer gets the new one, Team B would again gets a chance at the question. Since much of Infinite Rebounds is in protest of the injustices of the D&P method where this occurence is common and hated, it makes sense to prevent this as a special case and advance the first attempt marker to Team C. This would ensure that things aren't stuck at Team B's end.

Even if some do not agree with this point, it does make sense if one considers the spirit of the Infinite Rebounds philosophy. I speak from personal experience. In 2002, Swapnil did a quiz (called "NQuirer"?) at the BC. We got into this exact situation, and my team was in the position of Team B and would've benefited as we were not doing very well and needed every question we could get. Since this was quite a new situation (for some reason, no one at the BC had ever made a special note of this scenario), it wasn't quite clear who would get the next question, us or the team next to us. Many of the "learned figures", Harish and Niranjan quickly spotted the solution (the obvious one) i.e. we shouldn't get the chance. We (or rather me, since I was the senior guy in my team, so musn't pull Anup Mankar or Amrish into this), in a fit of silly gamesmanship, insisted that there was no rule or precedence governing this, and we should get it. I don't recall what was decided (most likely, it went against us by consensus, as I usually don't get my way at the BC ;-) ), but it did throw up a case for which we had no convention to govern. I will be the first to admit I was wrong that day and to avoid this situation in future, I like to make a special mention of it. That's why all these words are spent on it, so as to avoid acrimony in the use of such a splendid system as this.

For some incomprehensible reason, Infinite Rebounds isn't yet the universal standard for the conduct of general rounds of quizzes ("general" as in normal passing rounds that don't have any special funk such as Rapid Fire or Single Specialities etc.). It is inexplicable that even veteran quizhosts hang on to the D&P method for dear life. Either they are ignorant about it, or do not have the intellect to comprehend it (which must make it a pea-sized intellect), or they must be resisting its use. I don't claim to say that I can conclusively prove it is a fairer system than any other, but experience, personal and among the fraternity is overwhelmingly positive about it, or atleast in rejecting the conventional system. I also suspect some of the ancient quizhosts perceive a sense of having to give up their dignity and adopt a newer system; perhaps it would undermine all their previous quizzes. What else would explain one quiz organizer (a veteran quizzer himself) completely rejecting Infinite Rebounds and describing it as (paraphrased) "nonsense"? Unless he knew something we didn't. Would you hang on to a fast-eroding prestige by being stubborn or be fair to the participants? I think a great deal of quizzing inconsistencies are because these reputed quizmasters don't spend enough time thinking about their conduct of their quizzes and prefer keeping the show audience-centric rather than participant-centric. I don't buy the argument that Infinite Rebounds is hard to conduct; in fact it makes life easier. The usual points of confusion for someone conducting a quiz are how much does the team that just got the correct answer get in terms of points and who gets the next question. Both these are easy to answer in this format: everyone gets the same number of points each time and one just has to remember the last correct answer and/or the last starting point. As in many other walks of life, simple is elegant.

In an effort to try comparing and evaluating the efficacy of the two popular methods, Niranjan wrote a small simulation whose source code (in C) he posted on Inquizitive - apparently he's improved it since. It would make even more sense to run it with actual data from live quizzes rather than just test data, so some of us are trying to record actual sequences - I managed to do so from the last quiz I conducted, an internal office event called Pulse. I still haven't run the data against the sim. If anyone's interested in the code, just leave a comment behind here.

To summarise, I think since Infinite Rebounds is both widely accepted and perceived to be fair, its use must be encouraged all around. If somehow it can be introduced into the school quizzing circuit, it would expose kids to that at an earlier stage. Otherwise, it remains something of an elitist choice, but one that definitely exposes a badly organized quiz from the others. Of course, the most ideally fair way to hold a quiz is to allow everyone to attempt the same questions simultaneously and then the one who gets the most correct answers wins. This is obviously not possible in a "final-on-stage" situation. We can only get algorithms that approach that optimal solution. Infinite Rebounds is definitely here to stay as one such solution that is an excellent substitute for the practical conduct of a quiz.

Comments/Criticisms/Commendations/Contributions welcome as usual